Labour Dynamics in East Crete: Structural Characteristics and the Adoption of Sustainable Agricultural Practices ^†

Abstract

This study examines agricultural labour dynamics and sustainability practices in East Crete, assessing how labour structure, education, and input intensity shape ecological outcomes. Using data from 108 farms in Heraklion and Lassithi, we constructed composite indicators, such as Labour Intensity, Sustainability Engagement, and Training-Adjusted Labour indices. Analysis of 37 farms with data revealed a heterogeneous landscape. Traditional family-based systems persist alongside uneven shifts toward agroecological practices. The Training-Adjusted Labour Index correlated with reduced pesticide use, while subsidy participation alone was not a reliable predictor of sustainable behaviour. Findings highlight limits of compliance-based incentives and the importance of knowledge-driven transitions. This study advocates typology-informed policies and longitudinal research for future policy design.

1. Introduction

Mediterranean agriculture faces mounting environmental, demographic, and policy challenges [1]. In East Crete, traditional family-based labour persists, while sustainability imperatives like reduced chemical use have begun to reshape practices. However, little is known about how these transitions manifest at farm level. This study examines labour and input structures in East Crete and assesses how farms engage with agroecological principles. While incentives matter, prior studies stress the importance of knowledge and institutional context [2,3]. Regional studies often lack tools to capture these dynamics [4]. We address this gap by benchmarking farm-level indicators that reflect sustainability-relevant behaviours.

2. Materials and Methods

We analysed 2018 LIFT survey data from 108 farms in Heraklion and Lassithi, covering labour allocation, education, inputs, and participation in agri-environmental schemes [5]. From these data we derived seven policy-relevant indicators, grounded in the prior literature. The analysis focuses on 37 farms with complete records, applying descriptive statistics and value clustering to identify farm typologies and input-efficiency patterns. Although this is a subset, it retains sufficient variation to identify meaningful typologies. Given the 37-farm analytic sample, we did not conduct hypothesis tests or make causal claims; inference is descriptive/typological to inform targeted policy and provide preliminary evidence for future, larger-sample research.

3. Results and Discussion

The 37-farm sample spans diverse labour systems and sustainability strategies. Farmers averaged 26.8 years of experience, yet education was limited. Only 23.6% had post-secondary schooling, and 20% formal agricultural training.

Labour input varied widely. The Labour Intensity Index (labour hours per EUR output), ranged from 0.004 to 3.82 labour units per EUR of output, with a median of 0.10. This range reflects both semi-mechanized farms and manual, family-based labour, a pattern consistent with structural inertia in Mediterranean farming [6]. Input use also exhibited variability. Fuel intensity averaged EUR 0.37/output EUR (max 1.68). This range reflects unequal access to, or investment in mechanisation across farms. Fertiliser intensity had a mean of EUR 0.17/output EUR (max 0.88). These figures suggest that while some farms apply fertiliser sparingly or not at all, others maintain chemically intensive practices that are probably driven by yield expectations or agronomic tradition. The most notable variation was observed in pesticide intensity, which ranged from a near-zero value of 0.0011 to as high as EUR 2.94/output EUR. These ranges reveal unequal mechanization, varying reliance on chemical inputs, and contrasting pest management strategies. Together, they illustrate stark differences in production efficiency and environmental risk [7].

The Sustainability Engagement Index (SEI), constructed as a binary variable indicating whether a farm received agri-environmental payments, was positive for 92% of the sample. However, input intensities varied, showing that subsidies do not reliably predict ecological performance, echoing critiques of compliance-based frameworks [2]. To assess education effects, we constructed a Training Index (average of general and agricultural education), which averaged 0.22, with over half of farms scoring zero, underscoring major cognitive and technical gaps. We then developed the Training-Adjusted Labour Index, integrating education with environmental performance and adjusted for near-zero pesticide values. A safeguard constant of 0.01 was added to the denominator to avoid distortion from zero or near-zero pesticide values. Such a constant is a common practice in statistical analyses to maintain computational integrity. For a detailed discussion on the implications of adding constants to variables to handle zero values, see [8]. This index value ranged from 0 to 90.34, and farms in the upper quartile used significantly less pesticide. While descriptive, this supports the knowledge–practice link, suggesting that, where present, education does appear to contribute meaningfully to more ecologically conscious farm management [3,9].

Typology revealed two profiles: low-education/high-input farms with weak ecological performance despite subsidies and moderately educated/low-input farms with more coherent sustainability strategies [4]. The former constitutes a “training first” group (capability building before input reduction targets), the latter a “performance first” group (support linked to maintaining or advancing low-input outcomes). These contrasts suggest transitions depend more on internal capacity than external incentives.

4. Conclusions

Farming in East Crete combines traditional practices with signs of ecological adaptation. While most farms participate in subsidies, this does not consistently lead to sustainable behaviours. In contrast, farms with moderate training and lower input reliance align more closely with agroecological goals. As the analytic sample is small, these conclusions are not intended to be generalized beyond similar contexts; they serve as decision-support signals for policymakers. In addition, this study contributes to setting the groundwork for future studies in labour dynamics through the prism of sustainability, with larger samples.

The Training-Adjusted Labour Index reinforces the importance of education in reducing pesticide use, but the lack of formal training remains a key barrier to broader sustainability [3,9]. Future policy must move beyond one-size-fits-all incentives toward support that recognizes farm typologies and fosters peer learning [2,4]. Here, sustainability here is not imposed; it is cultivated through the same attentiveness farmers give to their land. To operationalize this, we suggest a two-track approach: build farmers’ capabilities where capacity is low (education, advisory, peer learning) and align incentives with verifiable low-input outcomes where capacity is adequate. Both tracks should be complemented by measures that strengthen market pull for sustainably produced goods.